Unlock instant, AI-driven research and patent intelligence for your innovation.

Efficient and robust high-speed neural networks for cell image classification

a high-speed, neural network technology, applied in the field of cell sorting, can solve the problems that all existing methods do not meet the speed requirements of typical applications including image activated cell sorting, and achieve the effects of high recall implementation, high accuracy, and high speed

Pending Publication Date: 2022-05-19
SONY CORP +1
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Benefits of technology

This patent describes a way to make neural networks that can classify cells based on their image. The method uses a difference between scores to determine if a cell is unclear or not. If it is, it is classified in a separate group, and if it is not, it is classified in the highest-scoring group. This method is fast, accurate, and efficient, and can be used for cell image classification.

Problems solved by technology

All of those existing methods do not meet the speed requirements of typical applications including image activated cell sorting.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Efficient and robust high-speed neural networks for cell image classification
  • Efficient and robust high-speed neural networks for cell image classification
  • Efficient and robust high-speed neural networks for cell image classification

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0015]A shallow two-convolutional-layer neural network architecture has been designed and established suitable for very high-speed real-time cell image classification. In addition, the algorithm is able to minimize the common trade-off pitfall between precision and recall to achieve high precision with high recall.

[0016]The neural networks with cell image classification address unmet needs of many cell image classification applications—mainly two key requirements: high accuracy and high speed. The existing classical machine learning-based method (e.g., boosting, svm with manually engineered features) and deep learning-based method cannot met the speed requirements. The method described herein not only runs very fast but also achieves high accuracy because of the shallow neural network architecture designed by architecture search techniques and the novel precision-recall trade-off module.

[0017]The neural networks with cell image classification are able to be used in image-based flow ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

An efficient and robust high-speed neural networks for cell image classification is described herein. The neural networks for cell image classification utilize a difference between soft-max scores to determine if a cell is ambiguous or not. If the cell is ambiguous, then the class is classified in a pseudo class, and if the cell is not ambiguous, the cell is classified in the class corresponding to the highest class score. The neural networks for cell image classification enable a high speed, high accuracy and high recall implementation.

Description

CROSS-REFERENCE TO RELATED APPLICATION(S)[0001]This application claims priority under 35 U.S.C. § 119(e) of the U.S. Provisional Patent Application Ser. No. 63 / 116,088, filed Nov. 19, 2020 and titled, “EFFICIENT AND ROBUST HIGH-SPEED NEURAL NETWORKS FOR CELL IMAGE CLASSIFICATION,” which is hereby incorporated by reference in its entirety for all purposes.FIELD OF THE INVENTION[0002]The present invention relates to cell sorting. More specifically, the present invention relates to image based cell sorting.BACKGROUND OF THE INVENTION[0003]Cell image analysis plays a more and more important role in biological and medical study, but existing methods cannot be utilized for real-time high speed applications, such as cell sorting. A deep neural network is usually built to achieve high accuracy classification; however, its complexity limits its usage in real-time applications that require very high speed.SUMMARY OF THE INVENTION[0004]An efficient and robust high-speed neural networks for cel...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06K9/00G06K9/62G06N3/04
CPCG06K9/00147G06K9/6227G06K9/6253G06K2209/05G06K9/628G06N3/04G06K9/00134G06V20/69G06V2201/03G06V10/764G06N3/08G06N3/048G06N3/045G06F18/2431G06V20/698G06V20/693G06F18/40G06F18/285
Inventor GONG, LIYULIU, MING-CHANG
Owner SONY CORP